2

I've run into an issue while recording audio with MXL-990 using Behringer U-Phoria UMC204HD.

We're recording audio with the above devices and video+audio with a video camera. The audio does not sync up with the video. It starts OK, but 10minutes in the audio comes in too fast - the video seems to lag behind. In other words: the sounds happen sooner than they should.

We've tried multiple cameras and multiple PCs for recording (linux, MacOSX), as well as multiple pieces of software (Audacity, Ardour, GarageBand, Oceanaudio), all yielding similar results. Changing the sample size/bitrate didn't help.

Here's an example of that. In the top track is the audio from the Behringer, and on the bottom track there's audio extracted from one of the cameras we've tested.

beginning of the recording

10 minutes in

There are two claps in the recording, 10 minutes apart. The first one is in sync, the other one, not so much. You can see the discrepancy in the screenshot above

Could it be that our interface is faulty? How can we confirm that?

  • Does the camera have a setting for variable frame rate and constant frame rate? (VFR/CFR) other users have found that to be an issue. – Timinycricket Nov 24 '19 at 22:05
  • @Timinycricket we've tested with 3 total different cameras on the same time, please look at this timeline: community.musictribe.com/vdgmh27479/attachments/vdgmh27479/… – Arkadiusz Wieczorek Nov 25 '19 at 8:34
  • Also be aware that the drift is locked in during the recording. Once you have made the recording, everything after that point is viewed independent of the clock in the audio interface. – Mark Nov 25 '19 at 8:50
1

The most likely culprit is that you are using a non-integer frame-rate with your camera, but when you are bringing the footage into the computer it is finding it's way into the timeline at a different frame-rate. Check the frame-rate of the camera. Sometimes the frame-rate is not what it seems.

| improve this answer | |
  • 1
    A lot of these cameras will tell you it's 30fps but in reality it's 29.97 non-integer. Feel free to dropbox me some of your footage and audio. mark.p.edwards@gmail.com – Mark Nov 24 '19 at 12:05
  • 1
    if you drop a 29.97fps footage into a 30fps timeline, your audio will slip for sure. – Mark Nov 24 '19 at 12:06
  • 1
    if it was an integer/non-integer related issue, the slip would be greater than what you are seeing. Unfortunately I think the problems is most likely a clocking issue with your recording device. The sampling clock isn't the same as the sampling clock running in the camera. Most professional equipment will permit 'genlock' to be used which allows you to clock both the recording device and the video device from the same source. Unfortunately with the devices you are using this is not possible. You will have to stretch the audio in post to match. – Mark Nov 24 '19 at 13:15
  • 1
    very unlikely that it is a faulty interface. the clock drift is less than 0.5% so this would not be considered a fault. – Mark Nov 24 '19 at 23:48
  • 1
    I think your only option is to apply time-stretch to the footage in order to align it with the footage from your cameras. You should use a slate at the start and end if you are going to be working with long takes. – Mark Nov 25 '19 at 7:41
0

This is a troubleshooting suggestion rather than an answer. StackExchange says that I don't have enough reputation points to post a comment.

Play a video on your TV and film it with one of your cameras while also recording it with your Behringer. Then, play back the video along with the two recordings, and see which recording aligns with the video.

If the problem is the Behringer, then there might be a way to calibrate its clock. If the problem is the camera, then you just need a more honest camera.

| improve this answer | |
  • Thank you for your suggestion. We will try that! If it turns out to be an issue with the behringer's clock, do you know how it can be adjusted? – Kuba Orlik Nov 26 '19 at 10:10
  • 1
    It can't be adjusted. – Mark Nov 26 '19 at 12:32
  • Talk to their tech support. For the accuracy that they advertise, I assume they use a crystal oscillator (as opposed to an LC circuit), and the crystal might be defective. It is possible to use an adjustable cycle counter for the clock, but that would have to be designed in. If they did implement that in their design, then their repair techs are probably the only ones who know the procedure. If the clock cannot be calibrated, then the whole circuit board would need to be replaced, and it's probably the main circuit board. – Arnold Cross Nov 26 '19 at 17:15

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.